# @empiricalrun/test-gen ## 0.81.2 ### Patch Changes - Updated dependencies [8b1be9e] - @empiricalrun/reporter@0.29.1 - @empiricalrun/test-run@0.17.1 ## 0.81.1 ### Patch Changes - Updated dependencies [76d0872] - @empiricalrun/shared-types@0.14.0 - @empiricalrun/llm@0.27.0 - @empiricalrun/cua@0.4.1 - @empiricalrun/dashboard-client@0.3.0 - @empiricalrun/reporter@0.29.0 - @empiricalrun/test-run@0.17.0 - videostil@0.3.6 ## 0.81.0 ### Minor Changes - 6fb8170: feat: playwright utils upgraded to use new core ### Patch Changes - Updated dependencies [6fb8170] - Updated dependencies [aa78c40] - @empiricalrun/cua@0.4.0 - @empiricalrun/reporter@0.29.0 - @empiricalrun/shared-types@0.13.0 - @empiricalrun/test-run@0.17.0 - @empiricalrun/dashboard-client@0.3.0 - @empiricalrun/llm@0.26.0 - videostil@0.3.5 ## 0.80.3 ### Patch Changes - Updated dependencies [12b14e5] - @empiricalrun/dashboard-client@0.3.0 ## 0.80.2 ### Patch Changes - Updated dependencies [2d6f146] - @empiricalrun/dashboard-client@0.2.1 ## 0.80.1 ### Patch Changes - Updated dependencies [3a315aa] - @empiricalrun/reporter@0.28.1 - @empiricalrun/test-run@0.16.1 ## 0.80.0 ### Minor Changes - ab403b8: fix: move json report location to avoid accidental agent commits ### Patch Changes - Updated dependencies [ab403b8] - @empiricalrun/test-run@0.16.0 ## 0.79.7 ### Patch Changes - Updated dependencies [c815fc2] - @empiricalrun/reporter@0.28.0 - @empiricalrun/test-run@0.15.0 ## 0.79.6 ### Patch Changes - Updated dependencies [b4e2fff] - @empiricalrun/cua@0.3.0 - @empiricalrun/llm@0.26.0 - videostil@0.3.5 ## 0.79.5 ### Patch Changes - b58777e: refactor: extract CUA agent into standalone @empiricalrun/cua package - fda9bc3: feat: extract DashboardAPIClient into standalone @empiricalrun/dashboard-client package - Updated dependencies [b58777e] - Updated dependencies [fda9bc3] - @empiricalrun/cua@0.2.0 - @empiricalrun/dashboard-client@0.2.0 ## 0.79.4 ### Patch Changes - dc2e6c9: feat: dashboard api client upgrade ## 0.79.3 ### Patch Changes - Updated dependencies [92254ba] - @empiricalrun/r2-uploader@0.9.1 - @empiricalrun/test-run@0.14.2 ## 0.79.2 ### Patch Changes - @empiricalrun/test-run@0.14.1 ## 0.79.1 ### Patch Changes - 48126ea: feat: cost tracking for overlay dismissals - Updated dependencies [48126ea] - @empiricalrun/llm@0.25.2 - videostil@0.3.4 ## 0.79.0 ### Minor Changes - 0dfc150: fix: exports config between packages ### Patch Changes - Updated dependencies [0dfc150] - @empiricalrun/test-run@0.14.0 ## 0.78.7 ### Patch Changes - Updated dependencies [1967ac3] - @empiricalrun/r2-uploader@0.9.0 - @empiricalrun/test-run@0.13.2 ## 0.78.6 ### Patch Changes - Updated dependencies [3ee1aec] - @empiricalrun/r2-uploader@0.8.0 - @empiricalrun/test-run@0.13.1 ## 0.78.5 ### Patch Changes - Updated dependencies [d270c6d] - Updated dependencies [2d9919d] - @empiricalrun/test-run@0.13.0 - @empiricalrun/r2-uploader@0.7.0 ## 0.78.4 ### Patch Changes - Updated dependencies [79a4e0f] - @empiricalrun/r2-uploader@0.6.0 - @empiricalrun/test-run@0.12.0 ## 0.78.3 ### Patch Changes - Updated dependencies [a94a7f7] - @empiricalrun/r2-uploader@0.5.0 ## 0.78.2 ### Patch Changes - Updated dependencies [25962dc] - @empiricalrun/llm@0.25.1 - videostil@0.3.3 ## 0.78.1 ### Patch Changes - 29eaffb: feat: added videostil as internal package - Updated dependencies [2187966] - Updated dependencies [897486a] - Updated dependencies [758c41a] - videostil@0.3.2 - @empiricalrun/llm@0.25.0 ## 0.78.0 ### Minor Changes - b01be05: feat: move task queue to r2-uploader package - e65313e: refactor: remove unused functions and streamline video analysis process ### Patch Changes - Updated dependencies [b01be05] - Updated dependencies [e65313e] - @empiricalrun/r2-uploader@0.4.0 - @empiricalrun/llm@0.24.0 ## 0.77.0 ### Minor Changes - bc38c65: feat: diff tab ui update with code review comments - e5e45e0: feat: code review agent output format update to v2 - 44373a4: feat: code review agent system prompt and ui update - 8a206cf: feat: optional line annotated diff on /diff endpoint for code review agent - 151dc8f: feat: fetchVideoAnalysis uses VideoAnalysisAgent under the hood - 49b44a6: feat: cr agent auto refresh ui with fresh code review data - 82a6fe7: refactor: chat model update in base agent v2 - f9a7023: feat: tool exec v2 integration with base agent v2 - a25735a: feat: added trace in code review agent ### Patch Changes - a45714d: fix: updated prompt to ask agent to format code description in markdown - ba65440: feat: convert image url to base64 uri before sending to llm - cf991ee: chore: file view tool for code review agent - eda0ccd: feat: support tool response image parts with urls - 1f7e8b0: chore: diff endpoint now also returns pr metadata, code review agent receives pr metadata - 42e19ee: fix: removed duplicate trigger on create pr tool for code review - 02c3532: fix: graceful handling of missing verdict in xml - infer verdict from cr agent line comments - 8f3a25f: feat: versioned code review responses - bd61b13: feat: change default model to gpt-5 - Updated dependencies [8ff49e5] - Updated dependencies [ba65440] - Updated dependencies [eda0ccd] - Updated dependencies [25bd9b6] - Updated dependencies [d56954c] - @empiricalrun/llm@0.23.0 - @empiricalrun/test-run@0.11.1 ## 0.76.0 ### Minor Changes - 9c9d6e5: feat: code review basic agent - ee8ce00: feat: fetchVideoAnalysis multi modal tool response - 86a1549: feat: code review agent v1 with workflow as an ad hoc tool call - 5339d2c: feat: VideoAnalysisAgent working in CLI with extractFrames tool ### Patch Changes - a6aa6d8: fix: bug where video analysis was set to empty string - @empiricalrun/llm@0.22.0 - @empiricalrun/test-run@0.11.1 ## 0.75.0 ### Minor Changes - dde3517: feat: trace zip req & res body update from network trace - 8668785: feat: enabled params for video analysis tool ### Patch Changes - 8c1935f: chore: feature flag in fetchVideo tool, UI changes to accomodate analysis on click with params - ab1be01: chore: introduce agent class - 9f1cb66: chore: enable none as flag value, lint changes - 1a7f487: fix: truncating the request body before appending to req res content in trace zip tool - e3f7028: feat: add ff to allow video analysis in test-gen session (admin only) - Updated dependencies [2a06e9b] - Updated dependencies [9f1cb66] - Updated dependencies [9fb0ba1] - Updated dependencies [8668785] - @empiricalrun/llm@0.22.0 - @empiricalrun/test-run@0.11.1 - @empiricalrun/r2-uploader@0.3.11 ## 0.74.2 ### Patch Changes - Updated dependencies [cdc0995] - @empiricalrun/test-run@0.11.0 - @empiricalrun/llm@0.21.0 ## 0.74.1 ### Patch Changes - Updated dependencies [6f2aa95] - @empiricalrun/ast-parser@0.0.10 ## 0.74.0 ### Minor Changes - de7be1a: feat: added console traces to trace.zip tool - 5fc3b85: feat: add fetchLastSuccessfulTestRun tool & api endpoint - 8956b5c: refactor: simplify default model selection by delegating to a single function - 7c4057e: feat: Made the listIssues tool call inline - 17fa058: feat: add model mode, mode based tool & system prompt - b52ec03: feat: triage summary tool with slack messages - 3613b5f: feat: updated tool defs, runtTest, fetchLastSuccessfulTestRun, viewFailedTestRunReport, updated last-successful-run endpoint - c0c97cf: feat: add chunking and dedup fs to avoid memory issues, add LLM usage to toolResponse - a03391a: feat: upload unique frames to video-analysis bucket with video hash - b404912: feat: added trace.zip tool for network traces (based on existing approach from inngest) - 1d90369: feat: permalink support in issues, updated IDashboardClient to getBaseUrl - 4b45e85: feat: split upload b/w frames and summary to decrease wait - cbaf687: feat: Added auto option in CLI (empty --model results in default fallback) - 5a8af12: feat: tools array based on agent mode or env - ce54ca8: feat: video analysis working in CLI & UI minor cleanup - 565a820: refactor: converted PUT calls to PATCH in update issue api - 681aebb: fix: moving trace dot zip tool to not be inline so it gets executed on tool execute - 773d1f8: feat: issue queries & tools update with new test metadata fields - 6585338: feat: added inline viewTestRunReport tool for triage - c0931db: feat: add issue(s) list & add tools - 2718944: fix: trim viewTestRunReportTool result size by removing steps from results array - aa14384: feat: add updateIssue tool - c112011: feat: add reset-chat flag to cli - 50c05ff: feat: added send triage summary tool - 9635632: feat: issues with health stats - 8eeb121: fix: project id filter fix, use-triage in cli ### Patch Changes - 1600c5d: feat: improvements to tool call telemetry - 01b693e: feat: moved file view commands to inline exec - 63a7b56: chore: remove some default model indirection - af74c99: fix: rm codeframes, steps, snippet from the test viewTestRunReport output - 9f33a7a: chore: fetch r2 files with suffix, add created_at to summary - 04940d6: fix: fix chunking error in ffmpeg local client with error handling - 028d845: chore: Added more tools to the inline execution roster - 01b693e: fix: added branch name field to the GitHub API and minor refactors in the test-gen package - 035febd: feat: runTest tool is now a common tool - 7fecf0a: feat: enable streaming on openai chat models - 4f598eb: feat: add gpt-5 - 4c8249a: test: fix agent-worker tests for parallel execution - ad91c42: fix: allow insert_text in str_replace_editor tool for claude 4 - 68ce1d1: fix: fetchVideoAnalysis error handling, minor cleanup - 2de74d0: chore: bump openai to 5.x - 41816c3: chore: enabled inline tool execution for fetch diagnosis. - 5651643: feat: recorder can pause before failing line when given a report url - 3d738f4: chore: split text editor tool call telemetry for commands - de01455: fix: moved project filter before inngest event send, added project id to log, separate test-gen utils exports - 40ef060: fix: support dir paths in inline file view - 5da0a60: feat: internal openai model supports multi-modal tool responses - 6d80d21: feat: video analysis review UI under admin route - fdc2ee8: feat: working video analysis with gemini-2.5-flash - 612e383: fix: parallel tool calling claude chat model - fe95cd7: chore: update prompts for triage session - 4b898bd: fix: elevate usage summary to agent loop - 7b3c12b: chore: simplify agent loop interface - 91d856c: fix: mode in agent-loop mandatroy, and being used - 5c91897: chore: Completed TODOs and refactored the inline tool identification logic" - ac01228: chore: upgrade to eslint 9 - Updated dependencies [cb503a6] - Updated dependencies [29eb6fb] - Updated dependencies [a6e6141] - Updated dependencies [63271bd] - Updated dependencies [9f33a7a] - Updated dependencies [c0c97cf] - Updated dependencies [7fecf0a] - Updated dependencies [4f598eb] - Updated dependencies [ad91c42] - Updated dependencies [ce54ca8] - Updated dependencies [2de74d0] - Updated dependencies [5651643] - Updated dependencies [5da0a60] - Updated dependencies [fdc2ee8] - Updated dependencies [612e383] - Updated dependencies [4b898bd] - Updated dependencies [89858b4] - Updated dependencies [f83a0b0] - Updated dependencies [7b3c12b] - Updated dependencies [98853ca] - @empiricalrun/llm@0.21.0 - @empiricalrun/r2-uploader@0.3.10 - @empiricalrun/ast-parser@0.0.9 ## 0.73.1 ### Patch Changes - a9b79d2: feat: auto-capture screenshots on test runs - a1b9060: feat: add interleaved thinking to claude - Updated dependencies [a1b9060] - @empiricalrun/llm@0.20.2 ## 0.73.0 ### Minor Changes - 81be84b: feat: use origin url for dashboard requests - b15ecf8: feat: using hash to generate sqs url for preview env - 0d3ea8d: feat: add fetchImageTool, update image check in claude - d5e16dc: refactor: unified dashboard api client ### Patch Changes - 643e859: fix: image type check in llm package, fetchImageTool description updated, image rendering fix in message component - d5a5ec7: fix: support artifacts with urls from json report - f8a4374: feat: add delete file tool, some clean up - 6423dcb: chore: cleaner tools API for chat agent loop - Updated dependencies [643e859] - Updated dependencies [0d3ea8d] - @empiricalrun/llm@0.20.1 ## 0.72.0 ### Minor Changes - 6ac65ed: feat: add custom sqs client with bug fix and error boundary - 23708d1: feat: add agent-worker app to run chat agent over durable objects ### Patch Changes - 6b1d98c: feat: trigger impacted tests computation on tool-responses - b6a04f5: feat: use webocket to get realtime updates on session - 4cbc287: feat: add (de)compress layer in the worker message management - a96a03c: chore: follow-ups from agent-worker first land - 32905df: chore: move dashboard methods from tool-execute-service to test-gen - 8e3c7a4: feat: working agent-worker for text messages - 5b0d43a: chore: logging tool call flow between agent-worker and dashboard - fc6f97c: fix: execSync handling for ubuntu and windows ci - 440e851: feat: extended thinking in claude - a23b38f: chore: add worker env in createChatModel & ToolCaller methods - 12c69cc: fix: error handling for non-retryable errors - 5ed01c4: fix: stop should reset askUserForInput state - e1d01c8: feat: add file info builder for github in test-gen - 450b79a: feat: show resolved projects for environments in cli - 7009d67: feat: show tool result preview in chat ui - d570c55: feat: add reporter function in agent worker - 622aa35: chore: split tool call service into caller and executor - 162e461: chore: file info clean up - 349003e: feat: working tool calls on agent-worker - d5c7696: test: add agent loop test harness - 7afa5c1: feat: better error handling & state updates, add langfuse trace id to worker session state - Updated dependencies [df226a5] - Updated dependencies [440e851] - Updated dependencies [a23b38f] - Updated dependencies [450b79a] - Updated dependencies [d5c7696] - @empiricalrun/llm@0.20.0 - @empiricalrun/test-run@0.10.9 ## 0.71.2 ### Patch Changes - d7f9025: chore: remove old session reporter which is no longer used - f44505a: fix: smaller responses from run-test tool - 001b324: fix: enforce codegen approach for overlay dismissals - Updated dependencies [f44505a] - @empiricalrun/test-run@0.10.8 - @empiricalrun/llm@0.19.4 ## 0.71.1 ### Patch Changes - ebafc76: chore: new commits to trigger publish - Updated dependencies [ebafc76] - @empiricalrun/llm@0.19.4 ## 0.71.0 ### Minor Changes - c2bcc63: feat: github cred to be always alive in cli ### Patch Changes - d326e62: feat: create draft request from cli - 6d5a283: fix: pw project resolution in dirs with spaces - fc5469c: fix: commit message duplicate skip ci - 64f4323: feat: tool call telemetry with posthog - Updated dependencies [6d5a283] - @empiricalrun/test-run@0.10.7 ## 0.70.2 ### Patch Changes - 2b0a5ec: feat: add session id to pr description - @empiricalrun/llm@0.19.3 ## 0.70.1 ### Patch Changes - 5149581: chore: update pr title to be dynamic ## 0.70.0 ### Minor Changes - 2daef2d: feat: upgrade-pkgs tool schema and PR details updated ### Patch Changes - 1348b1a: chore: refactor slack client into internal vs external usage - 4274dc5: fix: json report output should exist for video upload - 39b105b: fix: auto-merge checks in upgrade package tool ## 0.69.8 ### Patch Changes - a2ce1ec: fix: glass pane removal across browser context - ff7d9e6: fix: temp file computation for recorder - 5e9aeb7: fix: environments list command in test-gen cli - @empiricalrun/llm@0.19.3 ## 0.69.7 ### Patch Changes - bca8c8b: fix: codegen timing issue - 49a3382: feat: add commands for environments and download build - Updated dependencies [16f3927] - @empiricalrun/llm@0.19.3 ## 0.69.6 ### Patch Changes - c0c1a9a: fix: use gemini stable instead of preview - d67ca35: feat: add fetch-chat-session command, handle artifacts with same name - 9a923e2: chore: preview env resources are in ap-south-1 - Updated dependencies [c0c1a9a] - @empiricalrun/llm@0.19.2 ## 0.69.5 ### Patch Changes - 385cf7d: fix: ora needs dynamic import for nodejs 20 ## 0.69.4 ### Patch Changes - 5ded95d: fix: move open to dynamic import ## 0.69.3 ### Patch Changes - 298b5e5: feat: ask for project name in recorder flow - Updated dependencies [298b5e5] - @empiricalrun/test-run@0.10.6 ## 0.69.2 ### Patch Changes - e2f72b5: fix: handle video processing time and large uploads - Updated dependencies [e2f72b5] - @empiricalrun/llm@0.19.1 ## 0.69.1 ### Patch Changes - ff84083: feat: request verification for recording, more logs clean up - Updated dependencies [ff84083] - @empiricalrun/test-run@0.10.5 ## 0.69.0 ### Minor Changes - 64fa486: feat: user message can have video/webm attachments ### Patch Changes - a2cb0ab: feat: support video attachments for recorder requests - b533712: chore: remove verbose logs, hide pw glass pane - b9571e9: feat: cli command to clone and setup repo - Updated dependencies [a2cb0ab] - Updated dependencies [64fa486] - Updated dependencies [b533712] - @empiricalrun/llm@0.19.0 - @empiricalrun/test-run@0.10.4 ## 0.68.0 ### Minor Changes - cb957ea: feat: use cli user auth to upload recorder output ## 0.67.0 ### Minor Changes - 267b012: feat: add user login in test-gen cli ## 0.66.2 ### Patch Changes - 45dfd06: feat: show version comparison in banner - 7c66a47: feat: first working version of --use-recorder - 5bc45e1: feat: create requests from recorder cli - d0569de: fix: ensure browser agent is not stuck on page.pause - fd719f1: feat: upload recorder video and attach to request - a65a4d2: fix: path and ui issues ## 0.66.1 ### Patch Changes - 1a6bb19: chore: upgrade playwright in devDependencies - d3639c0: fix: update browsing agent artifact path for 1.53 - 881e856: fix: update system prompt to get PRs as output - cfb157a: fix: follow-ups for playwright reporter upgrade - 4062787: feat: update page.pause codegen to work with playwright 1.53 - Updated dependencies [cfb157a] - Updated dependencies [a533ee5] - Updated dependencies [4062787] - @empiricalrun/test-run@0.10.3 - @empiricalrun/llm@0.18.2 ## 0.66.0 ### Minor Changes - 576870a: feat: allow specifying package version in upgradePackages tool - e94a2da: feat: update upgradePackages tool to include input schema for package selection - dfb2007: feat: add upgradePackages tool ### Patch Changes - 33abadf: fix: text appendBranchNameToQueueUrl output is under 80 characters fixed - 1300a80: fix: branch name with dot should deploy tool execute service - 985d721: fix: git patch for new files - 6a19421: feat: collect git patch artifacts from text editor tools - 1a46013: feat: page.pause codegen works when tool execution is headed - 805f35f: fix: handle error when fetching environment variables in chat agent - 0a9ec78: feat: add o3 with reduced costs - c430cc0: fix: use mouse wheel to execute scroll in cua - Updated dependencies [9b873e3] - Updated dependencies [41c266d] - Updated dependencies [0a9ec78] - Updated dependencies [8a83b29] - @empiricalrun/test-run@0.10.2 - @empiricalrun/llm@0.18.1 ## 0.65.0 ### Minor Changes - 4623300: feat: added browser queue to add concurrency and avoid blocking of server due to browser dependent tools ### Patch Changes - 82acf53: feat: enable tools to contribute to cost calculations - 73dd841: feat: change getEnvironment tool to become listEnvironments - 8233d49: chore: remove ecs feature flag from Dashboard - 0b55884: feat: browser agent tool call returns images in tool result - 459d029: feat: pass feature flags to tool calls - 7712b2e: chore: move more types to shared-types package - 1b08d58: feat: tool response interface supports images for claude - 1b9087e: feat: improve feature flags ui, upgrade gemini-pro - Updated dependencies [82acf53] - Updated dependencies [1177d63] - Updated dependencies [7712b2e] - Updated dependencies [fb32af6] - Updated dependencies [1b08d58] - Updated dependencies [1b9087e] - @empiricalrun/llm@0.18.0 ## 0.64.3 ### Patch Changes - d8d624d: fix: pass mapped keys to codegen recordAction for proper code generation - addd52e: feat: add JSON value truncation to prevent oversized tool responses - f8a53b0: fix: PR description should not get concatenated across versions ## 0.64.2 ### Patch Changes - f05aac2: feat: add environment overrides for projects in tool calls - Updated dependencies [f05aac2] - @empiricalrun/test-run@0.10.1 - @empiricalrun/llm@0.17.3 ## 0.64.1 ### Patch Changes - 387f475: chore: move more types to shared-types package - 39f1aa8: fix: add exception handling to codegen - 09a1126: [WIP] refactor: process manager to have strict options to avoid failure due… - 1f74b60: fix: clarify how file create tool can create directories - e5f8110: fix: ripgrep should work with ampersand in query - b39d75c: feat: add build url to some tools, to run tests on preview urls - 99d8490: fix: override page screenshot to fix cua on mobile emulation - 226c361: fix: validate file existence in text edit tools, browser agent and run test tool and update fileName to filePath - 41fc766: feat: collect artifacts after process end, add fallback for actionSummary - d00c867: fix: add test for artifact extraction function and fix the bug - d018e6b: chore: change default model to claude sonnet 4 - b908890: feat: added useToolExecutionOnEcs feature flag to use ecs for tool execution - Updated dependencies [387f475] - Updated dependencies [587532f] - Updated dependencies [226c361] - Updated dependencies [0ef2e11] - Updated dependencies [8c7e9c6] - @empiricalrun/llm@0.17.3 - @empiricalrun/test-run@0.10.0 ## 0.64.0 ### Minor Changes - 1a5ec8d: feat: stateless tool execution and named args for tool execute fn ### Patch Changes - e5b9f7e: feat: consume repoPath in browser agent - 43dc453: fix: record locators before execution for more reliable codegen - 24d9415: fix: error message in browser agent tool call + removed unused code - Updated dependencies [1a5ec8d] - @empiricalrun/llm@0.17.2 ## 0.63.0 ### Minor Changes - f926e40: feat: consume repoPath in test run tool - 55c7913: feat: artifact collection is fire-and-forget at tool level - 2f1ee31: chore: consume repoPath and apikey in tools requiring Dashboard api call ### Patch Changes - c491cdd: fix: insert_line description in text editor tools - 0fea8f1: test: add a multiline old_str replacement test - 115a023: chore: allow grep tool to consume repoPath for the scope of search - 6c7740b: feat: consume repoPath for all file edit tools - Updated dependencies [f926e40] - Updated dependencies [2f1ee31] - @empiricalrun/test-run@0.9.4 - @empiricalrun/llm@0.17.1 ## 0.62.0 ### Minor Changes - f137da5: feat: collect artifact from browser agent, update Artifact UI. - e69b9d4: feat: implement artifact collection and enhance tool result handling in chat agent ### Patch Changes - 4923078: fix: add production=false to npm ls command - fe1ea6e: fix: debug logs for npm ls cmd, missing try-catch - 333b99f: fix: test run tool should report correct status - 6b6742c: refactor: create codegen abstraction for computer-use agent - a44e96c: feat: inject working directory and API key instead of deriving them from process context or env - 62fa1f2: feat: add support for claude 4 sonnet and opus - a0167b3: fix: add more validation in file creation tool - 696925d: feat: patch playwright for page.pause codegen approach - fc27187: chore: skip some flaky tests to ensure green reviews - 28f95d0: fix: ripgrep handles left brackets correctly - 1d39277: feat: implement ArtifactDisplay component to showcase artifacts in session details - fe3cdcb: Revert "feat: enhance tool execution with working directory and environment v…" - a44e96c: refactor: remove getTools method and initiate tool executors in constructor - c79ee3d: fix: add more debug logs for pm2/npm ls issue - Updated dependencies [333b99f] - Updated dependencies [3739bc0] - Updated dependencies [a44e96c] - Updated dependencies [62fa1f2] - Updated dependencies [e69b9d4] - Updated dependencies [fe3cdcb] - @empiricalrun/test-run@0.9.3 - @empiricalrun/llm@0.17.0 ## 0.61.0 ### Minor Changes - fca6729: feat: implement TypeScript compiler execution and error handling in strReplaceEditor ### Patch Changes - f33bdbf: fix: git changed line should cover untracked also - 1b4e3d0: fix: dont wait for fonts when agent needs screenshot - bbf735a: chore: remove unused test.only cleaner - b7a4821: fix: update pr tool prompt to encourage proactiveness - 779dd6e: feat: add email automation recipe to system prompt - 05f5184: fix: file create should create parent dir if required - Updated dependencies [d71508f] - Updated dependencies [79857b3] - Updated dependencies [fca6729] - @empiricalrun/test-run@0.9.2 - @empiricalrun/llm@0.16.1 ## 0.60.0 ### Minor Changes - 296c662: feat: implemented check for tool call existence and tool result non existence for tool execution - 255db12: feat: new linear client, Requests api routes, and fetch calls - 8544d9d: feat: ChatStateError integration in ChatState for better error handling & retry ### Patch Changes - b183b76: fix: remove tsx config reader in test-gen tool call - c4855ff: fix: file view tool does not read binary files - f821310: fix: ensure all text editor tools use 1-indexed line numbers - 06d3a2a: fix: skippped download build for failed build downloads - cff296d: fix: wrap tsx playwright config reader in a try-catch - 24d672a: feat: use playwright --list to get project names - Updated dependencies [24d672a] - @empiricalrun/test-run@0.9.1 - @empiricalrun/llm@0.16.0 ## 0.59.0 ### Minor Changes - e2b0318: feat: dashboard session migration from CLI working - 894c5b5: fix: stricter types for tool call args in canonical message - 919e7df: fix: session create, duplicate user message, add Model Options - 045e370: feat: stateless build download for tool execution - f2d5859: feat: canonical migration for claude and openai - e2b0318: feat: migrate dashboard chat sessions to canonical chat state - 4d3b46a: feat: migrate to new chat state version in cli run ### Patch Changes - 72f4577: feat: stop marking files as only and move to cmd builder approach - 82b3373: test: remove irrelevant or flaky tests - 3dbd4ff: fix: remove teardown skip since we interrupt agent with SIGINT - 7c77600: refactor: move test-gen file changes to prepare for removal - 562c25d: fix: update chat agent system prompt for using try/catch - 43656bf: refactor: split tool call execute into 2 methods - 51a9975: fix: unit tests for test-gen tool call - 772b93c: feat: gemini message transformer into canonical message type - 0c935ce: chore: chat model types are moved to shared-types pkg - 050fbd3: fix: model selection is ignored while creating a new session - ddfd009: fix: refactor chat agent and session list UI - 28682e2: feat: use test-run cmd builder in test-gen, remove test.only marking - 6b2ed6d: chore: use function args to tell if tool execution is remote - 511dbb1: test: skip tests for the old master agent - 7145e34: test: add new agent-loop test for openai canonical model - 43fb70f: feat: add retry functionality for chat sessions and update ChatAgent … - d1d1055: feat: integrate askUserForInput state into session handling - ef491c0: fix: incorrect build url getter - Updated dependencies [72f4577] - Updated dependencies [97cef16] - Updated dependencies [e2b0318] - Updated dependencies [772b93c] - Updated dependencies [894c5b5] - Updated dependencies [0c935ce] - Updated dependencies [919e7df] - Updated dependencies [045e370] - Updated dependencies [f2d5859] - Updated dependencies [919e7df] - Updated dependencies [ddfd009] - Updated dependencies [e2b0318] - Updated dependencies [28682e2] - Updated dependencies [b415e29] - Updated dependencies [09291a8] - Updated dependencies [aab69ae] - Updated dependencies [4d3b46a] - Updated dependencies [bc5996d] - @empiricalrun/test-run@0.9.0 - @empiricalrun/llm@0.16.0 ## 0.58.0 ### Minor Changes - 8c71fd1: feat: pass file info to system prompt to construct repo context ### Patch Changes - 637c97f: chore: moved model creation logic from test-gen to llm - Updated dependencies [0634233] - Updated dependencies [637c97f] - Updated dependencies [21b0768] - @empiricalrun/test-run@0.8.5 - @empiricalrun/llm@0.15.4 - @empiricalrun/r2-uploader@0.3.9 ## 0.57.2 ### Patch Changes - 8d69fe6: fix: use headless browser and authenticate project api key for all tool calls - d7c9b1d: fix: ripgrep tool is breaking on regex inputs - ca91d3c: fix: dont commit backup files in dashboard flow ## 0.57.1 ### Patch Changes - 1a0187f: fix: ripgrep searches should be case insensitive ## 0.57.0 ### Minor Changes - df30a88: feat: commit context to git on every tool file change ### Patch Changes - 8a847fd: feat: tracing for CUA when it runs in a tool - 597072c: feat: suggest unique occurences for old_str in text editor tools - 813995a: fix: grep tool should handle whitespaces in search string - Updated dependencies [629bbd2] - @empiricalrun/llm@0.15.3 ## 0.56.4 ### Patch Changes - 0d53865: fix: browser agent tool call has different exit code on linux - 8e5dcd2: feat: move to ripgrep for platform independent grep tool calls ## 0.56.3 ### Patch Changes - bf87a3a: fix: grep tool should not return error if no results are found - 607ec6d: fix: normalize old_str for newlines before running text editor tools - a7f2e24: fix: downgrade tsx logs to warn instead of errors - 1728128: feat: nesting of toolCall span under llm generation - Updated dependencies [a7f2e24] - @empiricalrun/test-run@0.8.4 ## 0.56.2 ### Patch Changes - d5828e1: feat: including toolCalling in langfuse trace - eb84f66: fix: Update environment variable handling - 40f6470: feat: added tool execution service - f49b645: feat: Add tool response API endpoint - Updated dependencies [f8cc82d] - Updated dependencies [f49b645] - @empiricalrun/llm@0.15.2 - @empiricalrun/test-run@0.8.3 ## 0.56.1 ### Patch Changes - d05bb69: chore: remove fs-extra - e1e2564: fix: avoid global promisify calls - 803f61f: fix: run type checks for file inserts in text editor tools - Updated dependencies [d05bb69] - @empiricalrun/test-run@0.8.2 ## 0.56.0 ### Minor Changes - b073084: feat: new APIs and refactor to introduce tool execute service ### Patch Changes - 93d7a0b: feat: add tool call for downloading builds - a58ac3f: feat: add dashboard agent integration and optimize tsx handling - Updated dependencies [87af227] - Updated dependencies [3831109] - Updated dependencies [a58ac3f] - Updated dependencies [f77e33d] - @empiricalrun/llm@0.15.1 - @empiricalrun/test-run@0.8.1 ## 0.55.0 ### Minor Changes - 508565d: feat: add support for openai chat model - 8da022c: feat: add environment fetching tool ### Patch Changes - 99e4e6e: refactor: decoupled agentloop and toolcall - Updated dependencies [cc4cb5e] - Updated dependencies [508565d] - @empiricalrun/llm@0.15.0 ## 0.54.1 ### Patch Changes - Updated dependencies [9c64161] - @empiricalrun/test-run@0.8.0 ## 0.54.0 ### Minor Changes - 8324aa6: chore: update auth header for dashboard requests ### Patch Changes - 769c3e7: feat: Implement environment-aware authentication with fallback - 4297ed5: chore: Move text editor tools from llm to test-gen package - 2ecdd64: fix: invalid tool calls should pipe back into the llm - a8f135e: fix: add package-lock.json to default exclude for grep tool - Updated dependencies [4297ed5] - @empiricalrun/llm@0.14.8 ## 0.53.13 ### Patch Changes - b847558: feat: cua codegen works for clicks in iframes - bb402a4: feat: accept framelocators in test-gen tool calls - aa38bee: feat: add valid playwright project names to system prompt ## 0.53.12 ### Patch Changes - 9edf60b: fix: remove check for modified files in create-pr tool - Updated dependencies [f25f7f1] - Updated dependencies [2b98129] - @empiricalrun/llm@0.14.7 ## 0.53.11 ### Patch Changes - fb78386: feat: enabled LLM tracing for chatAgent dashboard - Updated dependencies [a40b338] - @empiricalrun/llm@0.14.6 ## 0.53.10 ### Patch Changes - 088545c: feat: upload test reports from run-test tool calls - d003ea8: feat: add description to PRs created by chat agent - Updated dependencies [088545c] - @empiricalrun/test-run@0.7.7 ## 0.53.9 ### Patch Changes - 2e5d412: fix: grep tool should first return stdout, not stderr - 2619c58: fix: add selected model info to chat state - 58d170d: fix: skip all serial tests after browser agent is done ## 0.53.8 ### Patch Changes - ba5422f: fix: push intermediate file system changes to remote branch for chat agent ## 0.53.7 ### Patch Changes - 4f14f11: fix: handle git remote urls that have access tokens in them ## 0.53.6 ### Patch Changes - a3a1863: refactor: split chatagent into cli runner and agent loop - a32c076: feat: enabled LLM tracing for chat agent - eb89698: feat: used langfuse LLM tracing for claude and gemini usage - 9cc17cc: fix: import for chat state for dashboard - 17fcf83: feat: chat agent fetches and reports to the dashboard - 1c1fd00: feat: expose chatagent methods, starting with createChatState - c4c5a32: refactor: make chatmodels stateless and elevate state to chatagent - 48702e0: feat: checkout chat session branch before running chat agent - Updated dependencies [eb89698] - Updated dependencies [c4c5a32] - @empiricalrun/llm@0.14.5 ## 0.53.5 ### Patch Changes - 9f3cb10: feat: automated tracing for LLM call overlay dismiss - Updated dependencies [9f3cb10] - @empiricalrun/llm@0.14.4 ## 0.53.4 ### Patch Changes - 1426372: fix: remove stray console.log - 7efc3dc: feat: add page.goto to cua implementation + prompt edits - Updated dependencies [7efc3dc] - @empiricalrun/llm@0.14.3 ## 0.53.3 ### Patch Changes - 094b9f7: feat: add tool call for commit and push changes from chat agent - cc64ff1: feat: enable browser tool call to pick the right page to interact with ## 0.53.2 ### Patch Changes - Updated dependencies [0aeca9d] - @empiricalrun/llm@0.14.2 ## 0.53.1 ### Patch Changes - Updated dependencies [40fcbc2] - @empiricalrun/llm@0.14.1 ## 0.53.0 ### Minor Changes - d1d682d: feat: added typescript validation check in editor tools for gemini ### Patch Changes - Updated dependencies [d1d682d] - Updated dependencies [8533602] - Updated dependencies [e8035d6] - @empiricalrun/llm@0.14.0 ## 0.52.11 ### Patch Changes - ac2c8af: feat: made claude gemini handle parallel tool calling - Updated dependencies [ac2c8af] - @empiricalrun/llm@0.13.8 ## 0.52.10 ### Patch Changes - Updated dependencies [ae08b74] - @empiricalrun/llm@0.13.7 ## 0.52.9 ### Patch Changes - Updated dependencies [64d799d] - @empiricalrun/llm@0.13.6 ## 0.52.8 ### Patch Changes - 40b4619: fix: update browser agent prompt for todo prep - 7026a35: fix: missing package.json for cli launch ## 0.52.7 ### Patch Changes - b7115ea: feat: add ascii art banner to cli - 3d634c9: feat: added disk persistence for gemini - Updated dependencies [cebf4dc] - Updated dependencies [3d634c9] - @empiricalrun/llm@0.13.5 ## 0.52.6 ### Patch Changes - Updated dependencies [c7bba19] - @empiricalrun/llm@0.13.4 ## 0.52.5 ### Patch Changes - ea9ea2a: chore: upgrade gemini sdk and prune unused dependencies - Updated dependencies [ea9ea2a] - @empiricalrun/llm@0.13.3 ## 0.52.4 ### Patch Changes - 85665ef: feat: simplify text editor tool schemas for gemini - Updated dependencies [4f8137b] - Updated dependencies [85665ef] - @empiricalrun/llm@0.13.2 ## 0.52.3 ### Patch Changes - 6a19298: feat: changed gemini-2.5pro from exp to preview - cbe5823: fix: removed headed default from runTest tool schema and added function in role of toolRes Gemini - Updated dependencies [f4f4c5d] - Updated dependencies [6a19298] - Updated dependencies [cbe5823] - @empiricalrun/llm@0.13.1 ## 0.52.2 ### Patch Changes - c490603: feat: input initial prompt with markdown file - 68640d2: feat: handover from test-gen tool to chat agent with a summary message - ae91e37: fix: cap cua iterations, add tracing and improve logging - 0704b28: feat: zod schema for str_replace_editor for gemini to use this tool - 02a2439: feat: summarize actions done by cua and rename fileservice - 01fa143: feat: custom tool grep added for gemini - Updated dependencies [c490603] - Updated dependencies [486264f] - Updated dependencies [ae91e37] - Updated dependencies [0704b28] - Updated dependencies [3ed20a3] - Updated dependencies [01fa143] - @empiricalrun/llm@0.13.0 ## 0.52.1 ### Patch Changes - be00f0b: feat: update overlay dismissal to support multiple overlays at once ## 0.52.0 ### Minor Changes - a399a57: feat: added Gemini support to chat agent ### Patch Changes - 1b8d273: fix: clean up backup files - Updated dependencies [a399a57] - Updated dependencies [1b8d273] - Updated dependencies [99b0826] - @empiricalrun/llm@0.12.0 ## 0.51.6 ### Patch Changes - dc17737: fix: repo context dir not found - b058de5: feat: add app knowledge to chat agent system prompt - dbe2ace: fix: remove Anthropic type from chat agent - ce7fece: fix: improvements to cua usage in test-gen-browser tool call - b8b4eff: fix: crash when last-chat is not found - af84555: chore: remove reporter dependency from test-gen package - 2766be8: feat: introduce chat model interface to support multiple llms - Updated dependencies [dbe2ace] - Updated dependencies [b8b4eff] - Updated dependencies [2766be8] - @empiricalrun/llm@0.11.5 ## 0.51.5 ### Patch Changes - ae47eed: feat: add type checking to text editor tool call - beb704e: chore: move claude chat stuff to make space for gemini - 3d9c868: fix: system prompt for computer use agent - Updated dependencies [beb704e] - Updated dependencies [f283696] - @empiricalrun/llm@0.11.4 ## 0.51.4 ### Patch Changes - 7613f20: fix: update test-run-fetcher tool call response to include project - f57ab88: fix: add playwright best practices to system prompt - 83a4788: feat: use cua in chat agent tool call with feature flag - a8c11eb: fix: tool call log should show error - Updated dependencies [307a1cf] - @empiricalrun/llm@0.11.3 ## 0.51.3 ### Patch Changes - a32030e: feat: Formatted usage summary and coloured it - Updated dependencies [a32030e] - @empiricalrun/llm@0.11.2 ## 0.51.2 ### Patch Changes - ecd3c30: fix: show loader when claude is working - ad6c96f: fix: prune test-run tool call response for input tokens ## 0.51.1 ### Patch Changes - d04190f: fix: remove process.on listeners to avoid leaks - 75c7921: fix: show chat usage summary on ctrl+C - Updated dependencies [d04190f] - @empiricalrun/test-run@0.7.6 - @empiricalrun/llm@0.11.1 ## 0.51.0 ### Minor Changes - ac754ae: feat: enable disk persistence for chat state - 561aa8e: feat: add usage summary (tokens, cost) for chat agent ### Patch Changes - 3e3d937: fix: add some validations for browser agent tool call - Updated dependencies [ac754ae] - Updated dependencies [561aa8e] - @empiricalrun/llm@0.11.0 ## 0.50.4 ### Patch Changes - c907653: fix: Remove deprecated lodash.isequal in favor of node:util.isDeepStrictEqual - Updated dependencies [c907653] - @empiricalrun/test-run@0.7.5 ## 0.50.3 ### Patch Changes - aab7199: fix: uploading snapshots during test-gen - Updated dependencies [650733f] - @empiricalrun/llm@0.10.3 ## 0.50.2 ### Patch Changes - d808dda: feat: support claude3.5 as chat agent model with cli flag - 8e13e16: fix: move away from assets.empirical.run for test run reports - Updated dependencies [d808dda] - Updated dependencies [8e13e16] - @empiricalrun/llm@0.10.2 - @empiricalrun/reporter@0.23.2 ## 0.50.1 ### Patch Changes - b070af3: fix: error handling in test run tool - a94ef14: fix: chat agent system prompt for proactiveness - Updated dependencies [b070af3] - @empiricalrun/test-run@0.7.4 ## 0.50.0 ### Minor Changes - 0eeff70: feat: add test-run-fetcher tool call, response body changes ### Patch Changes - b14d5bf: feat: support headed executions of test run tool - Updated dependencies [b14d5bf] - @empiricalrun/test-run@0.7.3 ## 0.49.0 ### Minor Changes - 3d050ec: feat: Add grep tool for case-insensitive code search - ec8f23b: feat: Convert agent tools to use Zod schemas ## 0.48.1 ### Patch Changes - 13e3048: fix: add semi colon to terminate createTest line - b7092d0: fix: project detection should normalize paths - Updated dependencies [a2609f5] - @empiricalrun/llm@0.10.1 ## 0.48.0 ### Minor Changes - fc952c9: feat: Add new tool call to fetch diagnosis details ### Patch Changes - 32eaf6f: feat: add repo tree in ascii to chat system prompt - 6cc97ca: feat: use claude3.7 in chat agent for diagnosis auto-fix scenarios - 2e28c20: fix: master agent should not throw when used in tool call - f03ff97: feat: update master agent tool call to use TODO(agent) preparation method - ed273c2: fix: disable skills usage in master agent - 049102a: fix: support suites in agent tool calls - Updated dependencies [6cc97ca] - Updated dependencies [0fee9bf] - Updated dependencies [1f95e4b] - Updated dependencies [049102a] - @empiricalrun/llm@0.10.0 - @empiricalrun/test-run@0.7.2 ## 0.47.4 ### Patch Changes - 007ef65: feat: master agent tool call returns git patch on successful runs - c4ddc2c: feat: chat can use codegen agent as a tool call - e1903b7: feat: add master agent as a tool call for chat - 4b18733: chore: simplify interface to prepare for tool call - 6328aaf: feat: test-gen can use test-run package as a tool call - e4bf218: fix: update append-create-test prompt to emphasize task without location hints - 7748c7b: chore: remove options from prepare file workflow - 72a8cf2: feat: chat agent only needs prompt string - d995e4c: feat: enable conversation along with test-run tool call - 848cb88: fix: trace hierarchy in prep for master agent - Updated dependencies [72a8cf2] - Updated dependencies [242f659] - Updated dependencies [6328aaf] - Updated dependencies [d995e4c] - @empiricalrun/test-run@0.7.1 ## 0.47.3 ### Patch Changes - 56ed4eb: chore: remove parseJson utility, let it throw exceptions ## 0.47.2 ### Patch Changes - af97c0f: feat: cua agent can generate code - d7f1678: feat: support openai cua for overlay dismissal, bump openai to 4.87.3 - 09e880a: feat: add more actions for cua with better types - Updated dependencies [d7f1678] - @empiricalrun/llm@0.9.36 ## 0.47.1 ### Patch Changes - 1116d14: feat: run test-gen without requiring an installation step ## 0.47.0 ### Minor Changes - 9481edf: fix: import createTest from dist location instead of local package installation ## 0.46.11 ### Patch Changes - 0def0a2: feat: add inquirer prompts for missing CLI options ## 0.46.10 ### Patch Changes - b3ca74f: fix: bubble up browsing agent execution feedback for runtime planner - b484292: fix: pages summary for runtime planner works without scoped vars ## 0.46.9 ### Patch Changes - 6914e32: fix: use o3-mini to improve browsing agent accuracy - 22ef805: fix: improve accuracy for multiple pages in runtime planner ## 0.46.8 ### Patch Changes - 353ee4e: chore: move next action prompt to handlebars - 4c8ec02: chore: refactor action executor into a separate file - 05329d3: fix: skill action generator for multiple pages - 46b3e75: feat: add observation tool call to getNextAction - ef4842b: test: scroll and click inside a scrollable div element - ea5f6ef: chore: refactor skill execute action - 8ad1b4c: test: add a test for skill usage that creates a state variable - 1922ab3: chore: Refactor browser tests structure - 295b611: test: add test for skill usage and code gen ## 0.46.7 ### Patch Changes - f2e9d28: fix: button in class name does not mean clickable ## 0.46.6 ### Patch Changes - 8065639: fix: check for svg in descendants of provided html element and not direct children - f5f12f5: feat: Move icons knowledge to .empiricalrun directory ## 0.46.5 ### Patch Changes - cce1c90: fix: nesting for llm tracing ## 0.46.4 ### Patch Changes - 0c29798: test: fix assertion for icon registry test ## 0.46.3 ### Patch Changes - f26142f: fix: browsing agent giving wrong executed action - 112b429: test: enable parallel test execution in playwright ## 0.46.2 ### Patch Changes - 9770fd6: test: fill action with multiple pages - c723f61: fix: auto fix didnt work - e6d6174: chore: remove sentry from worker, service and test-gen ## 0.46.1 ### Patch Changes - 0aa2054: test: add failing test for annotation enrichment for iframes - 21f5534: fix: enrich annotations inside iframes ## 0.46.0 ### Minor Changes - b6dfc91: feat: create icon registry for test repositories ### Patch Changes - b6dfc91: feat: cache icon description generated through LLM ## 0.45.1 ### Patch Changes - 04db1a6: chore: better types for playwright action generators - 1b1815d: chore: move runtime planner to handlebars - 62800d7: fix: prevent sending test run events to sentry ## 0.45.0 ### Minor Changes - 698ad31: feat: add approval in skill execute ### Patch Changes - 23f875f: fix: abstract the communication method of human in loop ## 0.44.0 ### Minor Changes - c206ea8: feat: Add suites CLI flag for test generation ### Patch Changes - 57ea65d: fix: prompt for create test using code agent to honour the file name and test case name ## 0.43.3 ### Patch Changes - 0b4cffa: feat: add human in the loop for planner - 694e202: fix: report gen assets should not throw in cli ## 0.43.2 ### Patch Changes - 5334ba8: test: added failing test for multiple annotations for buttons - 385d2c6: fix: disable sentry for local development environment - 7b47902: chore: reduce assumptions of repo dir being process.cwd - 1d0a746: feat: new cli args for --name, --file, and --prompt - 0f342af: fix: break master agent loop if element is not visible after scroll ## 0.43.1 ### Patch Changes - 64f275a: feat: add `--token` arg name to cli interface ## 0.43.0 ### Minor Changes - 46c0dab: feat: autofix workflow v1 ### Patch Changes - Updated dependencies [46c0dab] - @empiricalrun/llm@0.9.35 ## 0.42.29 ### Patch Changes - f39500b: fix: removed explicit delays from master agent ## 0.42.28 ### Patch Changes - df9f428: feat: add distributed tracing for ci-service, ci-worker, test-gen and test-run ## 0.42.27 ### Patch Changes - 9fc1b40: test: increase timeout for lexical test - 1f18628: fix: separate out create test and repo edit LLM calls and remove str replace flag ## 0.42.26 ### Patch Changes - 070c16a: feat: config for master agent sanity test suite ## 0.42.25 ### Patch Changes - Updated dependencies [2babfe0] - @empiricalrun/llm@0.9.34 ## 0.42.24 ### Patch Changes - 005632b: fix: restore z-index after annotations - 4f8e042: chore: use handlebars compiler from llm package ## 0.42.23 ### Patch Changes - b11f2cc: feat: move handlebars compiler to llm package - 853384f: chore: Replace getPrompt with compilePrompt for colocated prompts - Updated dependencies [b11f2cc] - @empiricalrun/llm@0.9.33 ## 0.42.22 ### Patch Changes - f3f9abf: fix: Identifier 'MAX_Z_INDEX_FOR_DOM' has already been declared ## 0.42.21 ### Patch Changes - 7b19602: feat: use overlay text content to improve overlay dismissal accuracy - b6d1326: fix: z-index scaling for better overlay dismissals - f462463: chore: adopt shared-types package ## 0.42.20 ### Patch Changes - 9488f81: fix: infer-agent prompt template - b625749: chore: move infer-agent and planner prompts to handlebars - 5fb977c: test: clean up some unreliable tests - 8914542: feat: run planner inside master agent flow ## 0.42.19 ### Patch Changes - c36efe4: chore: remove any type for get-next-action output - ebb0bfa: feat: support images in handlebar prompts - 63ed479: fix: remove verification and looping inside browsing agent ## 0.42.18 ### Patch Changes - 6f876ea: fix: remove code generation feedback loop from repo edit - 658451e: fix: tracing for vitests - Updated dependencies [658451e] - @empiricalrun/llm@0.9.32 ## 0.42.17 ### Patch Changes - 65e821a: chore: remove dead code around selector hints - 286a1a5: fix: annotation test ## 0.42.16 ### Patch Changes - Updated dependencies [9a81634] - Updated dependencies [f2c3344] - @empiricalrun/reporter@0.23.1 ## 0.42.15 ### Patch Changes - f7b6045: fix: reduced border to 1px for annotation container ## 0.42.14 ### Patch Changes - da02df2: fix: added tool call for code generation and writing code updates to file ## 0.42.13 ### Patch Changes - 161b404: fix: added colors in annnotations ## 0.42.12 ### Patch Changes - Updated dependencies [c95a819] - @empiricalrun/reporter@0.23.0 ## 0.42.11 ### Patch Changes - 98a8417: fix: handle JSON parse with try and catch ## 0.42.10 ### Patch Changes - ad2e2c5: fix: annotations for small elements should be outside the container - 6471bf9: test: updated timeouts - fabdfd6: fix: moved lexical vars check to o3-mini and increased timeout - 61bc738: fix: added message for scroll on dashboard - Updated dependencies [fabdfd6] - @empiricalrun/llm@0.9.31 ## 0.42.9 ### Patch Changes - 7560d20: chore: fix test gen configs in tests - 13eb978: fix: tests for create-test-block and move prompt to hbs - 128123f: fix: added scroller method to master agent - 347e20f: feat: prompt builder with handlebars templates ## 0.42.8 ### Patch Changes - 18444bc: fix: pass action to BA in case of no annotations ## 0.42.7 ### Patch Changes - b6879bb: test: added assert scenario for annotations ## 0.42.6 ### Patch Changes - b0b578c: fix: added page for token debugging - 25ae4b6: fix: added preference in trace for annotations - 9247f19: fix: assertion capability with selector hints ## 0.42.5 ### Patch Changes - 8f5a315: fix: trace url is clickable - 02784a8: chore: move browser-injected-scripts to enable lint - Updated dependencies [4df759f] - Updated dependencies [1749342] - @empiricalrun/llm@0.9.30 ## 0.42.4 ### Patch Changes - 7ae0581: fix: adding userContext fixture caused syntax error ## 0.42.3 ### Patch Changes - 55a70ee: fix: revert to old version of playwright - Updated dependencies [55a70ee] - @empiricalrun/reporter@0.22.1 ## 0.42.2 ### Patch Changes - 87743d9: fix: fill action in case of multi page and context ## 0.42.1 ### Patch Changes - ce2b8c7: fix: filter allowed file extensions for repo agent input ## 0.42.0 ### Minor Changes - 86e5a4c: feat: merge generate requested change flow to test-gen ## 0.41.6 ### Patch Changes - cee101c: fix: search and replace code for repo edit agent ## 0.41.5 ### Patch Changes - da4d7f1: fix: added strict check for project repo name ## 0.41.4 ### Patch Changes - 399f968: test: fix eslint mock test output - 4c35992: fix: import statements with alias were getting removed from file ## 0.41.3 ### Patch Changes - 1d39892: fix: repo edit exception ## 0.41.2 ### Patch Changes - 8d6f8be: fix: enable type check in repo agent - 6573a58: fix: move generate requested change method to test-gen package - 54e7e03: fix: avoid uploading screenshots if reporter config is not set - 3eff6a8: fix: updated prompt for create test ## 0.41.1 ### Patch Changes - b93e004: chore: fix aws s3 sdk client version - Updated dependencies [b93e004] - @empiricalrun/r2-uploader@0.3.8 ## 0.41.0 ### Minor Changes - 02c37c0: feat: upgrade pw version ### Patch Changes - Updated dependencies [02c37c0] - @empiricalrun/reporter@0.22.0 ## 0.40.1 ### Patch Changes - b45c035: fix: sequentially execute file updates and validation in codegen - b3069f1: fix: handle empty files while applying prettier format ## 0.40.0 ### Minor Changes - 451c840: feat: pro 1165 adding interactible live stream of test generation ### Patch Changes - 7989789: fix: replaced implementation for create test with repo edit ## 0.39.0 ### Minor Changes - f351f94: Feature/test gen package real time updates ## 0.38.56 ### Patch Changes - 17dcddc: fix: added formatted output in trace for easier debugging ## 0.38.55 ### Patch Changes - e799273: fix: added preferences in annotation method ## 0.38.54 ### Patch Changes - 1041dc0: fix: run test-gen always in headed mode ## 0.38.53 ### Patch Changes - 1f87882: fix: remove selector hints flag from UI and enable it by default ## 0.38.52 ### Patch Changes - f4d894d: fix: remove playwright dependency from test-gen - Updated dependencies [f4d894d] - @empiricalrun/llm@0.9.29 ## 0.38.51 ### Patch Changes - a392f9c: chore: lint fixes - a9c4133: fix: remove environment variable from command ## 0.38.50 ### Patch Changes - 5826c61: fix: supporting streaming for headed mode for playwright tests ## 0.38.49 ### Patch Changes - a25f06d: fix: repo editor not able to make changes to the method ## 0.38.48 ### Patch Changes - 3864b3d: fix: typos in repo edit prompt ## 0.38.47 ### Patch Changes - 28a482d: fix: max_completion value for test update feedback agent ## 0.38.46 ### Patch Changes - 4ca5faa: fix: feedback for wrong oldCode in coding agent output ## 0.38.45 ### Patch Changes - efb88aa: fix: update fill action to use state variables properly ## 0.38.44 ### Patch Changes - e85633b: fix: parsing logic in appending scope variables in createTest ## 0.38.43 ### Patch Changes - 05f9a03: fix: apply file changes works without test case context ## 0.38.42 ### Patch Changes - 2e5073f: feat: add new session button ## 0.38.41 ### Patch Changes - eb46866: test: skip inspiration test that fails a lot ## 0.38.40 ### Patch Changes - Updated dependencies [d541939] - @empiricalrun/reporter@0.21.6 ## 0.38.39 ### Patch Changes - 7152cf7: fix: test gen invalid key issue ## 0.38.38 ### Patch Changes - ea74baf: feat: add support for repo editor agent - Updated dependencies [ea74baf] - @empiricalrun/reporter@0.21.5 ## 0.38.37 ### Patch Changes - e9daf03: chore: log line when update flow output is dropped - a552a6e: fix: convert trace url into hyperlink for coding agent ## 0.38.36 ### Patch Changes - c7c0cc7: feat: migrate from testcase to session for test generation ## 0.38.35 ### Patch Changes - 351950a: fix: allow gist generation for setup project test cases - 21d58e7: feat: add session status ## 0.38.34 ### Patch Changes - f3519a9: fix: maintain test alias in imports after master agent ## 0.38.33 ### Patch Changes - Updated dependencies [65061d8] - @empiricalrun/llm@0.9.28 ## 0.38.32 ### Patch Changes - efed142: fix: increase token limit ## 0.38.31 ### Patch Changes - 6060a78: fix: popup auto dismiss at runtime ## 0.38.30 ### Patch Changes - dbe713b: fix: auto dismiss popup at test runtime ## 0.38.29 ### Patch Changes - 32ef92e: fix: removed master agent v1 ## 0.38.28 ### Patch Changes - dc8d538: fix: ordering of npm install ## 0.38.27 ### Patch Changes - 931191c: fix: avoid enriching description of annotated elements if already provided ## 0.38.26 ### Patch Changes - 05b8b6c: feat: removed checks in canRunMasterAgent ## 0.38.25 ### Patch Changes - d56544d: fix: send the last screenshot after generation ## 0.38.24 ### Patch Changes - 20455e2: fix: segregate master agent and annotation prompts - 85c32a2: fix: failed action prompt ## 0.38.23 ### Patch Changes - 31a68b4: fix: downgraded claude to claude-3-5-sonnet-20240620 - Updated dependencies [31a68b4] - @empiricalrun/llm@0.9.27 ## 0.38.22 ### Patch Changes - 03be230: evals: add scores for append-create-test-block - d6c0198: fix: multiple append create test block - d6c0198: fix: add retry for agent if there are syntax issues ## 0.38.21 ### Patch Changes - Updated dependencies [33c6c28] - @empiricalrun/reporter@0.21.4 ## 0.38.20 ### Patch Changes - Updated dependencies [5231c86] - @empiricalrun/r2-uploader@0.3.7 ## 0.38.19 ### Patch Changes - ba654cf: fix: revert annotations ## 0.38.18 ### Patch Changes - 683da08: fix: replaced master-agent-action with master-agent-execute-action in trace - ebf4861: fix: hover tool call ## 0.38.17 ### Patch Changes - f13435d: fix: usage example retriever for skills ## 0.38.16 ### Patch Changes - 6163918: fix: security policy injection for locators ## 0.38.15 ### Patch Changes - 91ded8f: fix: incorrect annotations ## 0.38.14 ### Patch Changes - bd5c945: fix: run update scenario prompts on claude - 31f8805: fix: runtime planner calling out actions as done - Updated dependencies [bd5c945] - @empiricalrun/llm@0.9.26 ## 0.38.13 ### Patch Changes - babf00a: fix: updated check for instance element ## 0.38.12 ### Patch Changes - 3739ba6: fix: userContext not getting imported in case of edit test - 368a859: fix: skill wasn't used in case task didn't contain next action details - 3d0185e: fix: reduce payload for fix ts error call ## 0.38.11 ### Patch Changes - fb4c866: fix: parent and child elements getting highlighted ## 0.38.10 ### Patch Changes - e3f04e6: fix: add support for injecting script before agent action ## 0.38.9 ### Patch Changes - 134f282: fix: annotations position after scroll ## 0.38.8 ### Patch Changes - deb93bd: fix: scroll while annotating ## 0.38.7 ### Patch Changes - b13b57d: fix: remove duplicate code ## 0.38.6 ### Patch Changes - 72c09d9: feat: added support for annotations in iframe ## 0.38.5 ### Patch Changes - 50c7a9a: feat: add support for page variables and state variables ## 0.38.4 ### Patch Changes - be5370d: chore: send skill usage message on chat ## 0.38.3 ### Patch Changes - ff76b50: fix: return first visible element matching with the CSS selector ## 0.38.2 ### Patch Changes - 132c2b4: fix: generation assets failure shouldnt stop test generation - 213c357: fix: added available annotations in master agent prompt ## 0.38.1 ### Patch Changes - c2cf877: fix: get appropriate skill should return correct file path ## 0.38.0 ### Minor Changes - 2444057: feat: add support for planner agent ### Patch Changes - 96bc299: fix: extract page vars is failing - edce85d: fix: annotate elements where opacity is not set ## 0.37.4 ### Patch Changes - 46e6f61: test: added test for fetch pom skills - 5315b91: fix: remove support for page.reload action - 9b4894a: fix: identification of agent stuck in loop - 8c6bab2: fix: regex for extracting test steps suggestion - 0128934: fix: added code in trace for master agent ## 0.37.3 ### Patch Changes - da47955: fix: exports path in package.json ## 0.37.2 ### Patch Changes - Updated dependencies [3eb7547] - @empiricalrun/llm@0.9.25 ## 0.37.1 ### Patch Changes - 3526169: fix: updated prompt for master agent next step ## 0.37.0 ### Minor Changes - 97e5659: feat: add append create test agent evals ### Patch Changes - 2dfecf6: fix: infer agent before enrichment call - 5b4509b: updated annotation hints for vue ## 0.36.6 ### Patch Changes - 63f1398: fix: triggerHintsFlow trace span nesting - Updated dependencies [42b4490] - @empiricalrun/llm@0.9.24 ## 0.36.5 ### Patch Changes - 32f1cb0: fix: do not skip teardown if test-gen is triggered for it ## 0.36.4 ### Patch Changes - c40ac5b: fix: remove group from test-gen-token ## 0.36.3 ### Patch Changes - 7e635e3: fix: remove testGroup usage inside test-gen and reporter - Updated dependencies [7e635e3] - @empiricalrun/reporter@0.21.3 ## 0.36.2 ### Patch Changes - 8c238de: fix: add trace for triggerHintsFlow ## 0.36.1 ### Patch Changes - e9a4eb5: fix: send file_path test-gen, ci-service, and ci-worker ## 0.36.0 ### Minor Changes - 89f9bdd: feat: add support for context variables ### Patch Changes - ba10561: fix: consider "setup" and "teardown" aliases as test block - Updated dependencies [a2eb6a1] - @empiricalrun/llm@0.9.23 ## 0.35.9 ### Patch Changes - 04ca7f3: feat: add locator.press action ## 0.35.8 ### Patch Changes - 461f3dd: fix: add logs for generation stopped due to stop session requested ## 0.35.7 ### Patch Changes - 2145207: fix: vitest config to exclude browser tests ## 0.35.6 ### Patch Changes - 948f534: fix: update verification agent prompt ## 0.35.5 ### Patch Changes - 8e18e5b: feat: add scenario code agent evals - d6f9de2: fix: add tests for annotation script ## 0.35.4 ### Patch Changes - 4bddc82: fix: annotate all non-blocked elements, outside the viewport too ## 0.35.3 ### Patch Changes - 19f5344: fix: added test url in trace metadata ## 0.35.2 ### Patch Changes - 2821db3: fix: handle agent getting stuck in loop - 0fe1f6e: fix: load annotate-elements script from disk ## 0.35.1 ### Patch Changes - Updated dependencies [557324f] - @empiricalrun/llm@0.9.22 ## 0.35.0 ### Minor Changes - 069347f: feat: add support for master agent evals - 11e4cbd: feat: add fetch skills agent evals ### Patch Changes - 297508d: fix: langfuse key errors - Updated dependencies [069347f] - Updated dependencies [297508d] - @empiricalrun/llm@0.9.21 ## 0.34.5 ### Patch Changes - dc5718a: feat: add support for evals - Updated dependencies [06cf0d8] - @empiricalrun/llm@0.9.20 ## 0.34.4 ### Patch Changes - Updated dependencies [2dafa69] - @empiricalrun/llm@0.9.19 ## 0.34.3 ### Patch Changes - Updated dependencies [53aabd4] - @empiricalrun/llm@0.9.18 ## 0.34.2 ### Patch Changes - 3b04b0e: chore: stop using static env vars - Updated dependencies [3b04b0e] - @empiricalrun/llm@0.9.17 ## 0.34.1 ### Patch Changes - Updated dependencies [25f3601] - @empiricalrun/llm@0.9.16 ## 0.34.0 ### Minor Changes - b0c5c81: feat: expose hints workflow behind feature flag ## 0.33.16 ### Patch Changes - 696d17b: fix: incorrect agent picked for the generation - e4558cf: chore: make trace url clickable for coding agent ## 0.33.15 ### Patch Changes - 4091222: feat: infer master / code agent from the task provided ## 0.33.14 ### Patch Changes - ec95c2a: fix: update flow prompt ## 0.33.13 ### Patch Changes - Updated dependencies [76e979c] - @empiricalrun/llm@0.9.15 ## 0.33.12 ### Patch Changes - Updated dependencies [398fa34] - Updated dependencies [b1343f9] - @empiricalrun/reporter@0.21.2 - @empiricalrun/llm@0.9.14 ## 0.33.11 ### Patch Changes - e6bbf04: fix: stop button is not working during generation ## 0.33.10 ### Patch Changes - Updated dependencies [04be13c] - @empiricalrun/llm@0.9.13 ## 0.33.9 ### Patch Changes - 1ef7d96: fix: check for lint package ## 0.33.8 ### Patch Changes - Updated dependencies [e547617] - @empiricalrun/llm@0.9.12 ## 0.33.7 ### Patch Changes - Updated dependencies [a7d11d2] - @empiricalrun/llm@0.9.11 ## 0.33.6 ### Patch Changes - Updated dependencies [ff11d3c] - @empiricalrun/llm@0.9.10 ## 0.33.5 ### Patch Changes - b48498a: fix: generation details missing in auto-fix flow ## 0.33.4 ### Patch Changes - Updated dependencies [a1e8599] - @empiricalrun/r2-uploader@0.3.6 ## 0.33.3 ### Patch Changes - Updated dependencies [5901db5] - @empiricalrun/r2-uploader@0.3.5 ## 0.33.2 ### Patch Changes - Updated dependencies [604f9a4] - @empiricalrun/reporter@0.21.1 ## 0.33.1 ### Patch Changes - 48e5628: fix: skip teardown files during test-gen ## 0.33.0 ### Minor Changes - 61aa7b7: feat: collect and update generation table with updates ### Patch Changes - Updated dependencies [61aa7b7] - @empiricalrun/reporter@0.21.0 ## 0.32.8 ### Patch Changes - Updated dependencies [b5379b8] - @empiricalrun/llm@0.9.9 ## 0.32.7 ### Patch Changes - Updated dependencies [e33656e] - @empiricalrun/llm@0.9.8 ## 0.32.6 ### Patch Changes - Updated dependencies [cf1740a] - @empiricalrun/llm@0.9.7 ## 0.32.5 ### Patch Changes - 4e0b694: fix: agent code not getting updated ## 0.32.4 ### Patch Changes - Updated dependencies [e20abfb] - @empiricalrun/llm@0.9.6 ## 0.32.3 ### Patch Changes - 0435eed: fix: geenration error incorrectly shown ## 0.32.2 ### Patch Changes - d2bc4c5: fix: add logs and traces - Updated dependencies [0eb7700] - Updated dependencies [55f21cf] - @empiricalrun/llm@0.9.5 ## 0.32.1 ### Patch Changes - 7373dea: fix: project detection should use environment pw project filter - Updated dependencies [ff77220] - @empiricalrun/r2-uploader@0.3.4 ## 0.32.0 ### Minor Changes - e22c1eb: feat: add support for extracting lexical scoped variables ## 0.31.26 ### Patch Changes - 9cfa0b9: feat: add support for user context in fixtures ## 0.31.25 ### Patch Changes - Updated dependencies [9d17374] - Updated dependencies [d78a90f] - @empiricalrun/reporter@0.20.7 ## 0.31.24 ### Patch Changes - Updated dependencies [994f76c] - @empiricalrun/reporter@0.20.6 ## 0.31.23 ### Patch Changes - Updated dependencies [0c5e656] - @empiricalrun/reporter@0.20.5 ## 0.31.22 ### Patch Changes - c00fd01: fix: move the reason to top in the function calls - Updated dependencies [6b596c8] - @empiricalrun/reporter@0.20.4 ## 0.31.21 ### Patch Changes - Updated dependencies [21190fa] - @empiricalrun/reporter@0.20.3 ## 0.31.20 ### Patch Changes - fcacf06: feat: add support for poms in master agent - Updated dependencies [fcacf06] - @empiricalrun/llm@0.9.4 ## 0.31.19 ### Patch Changes - b097c97: feat: add hover action on browsing agent ## 0.31.18 ### Patch Changes - 3600f40: fix: pass suites in prepareFileForMasterAgent flow ## 0.31.17 ### Patch Changes - c58e91c: fix: show error message and stack in test-gen failure ## 0.31.16 ### Patch Changes - ad4af33: fix: update test agent halucinating in larger context ## 0.31.15 ### Patch Changes - 71898fe: fix: handle json string responses of o1 ## 0.31.14 ### Patch Changes - 89226da: fix: xpath leading to root element ## 0.31.13 ### Patch Changes - 759bfc0: fix: accuracy of element selector ## 0.31.12 ### Patch Changes - a8f2ea6: chore: add test for master agent - Updated dependencies [10f554d] - @empiricalrun/reporter@0.20.2 ## 0.31.11 ### Patch Changes - Updated dependencies [9623b74] - @empiricalrun/reporter@0.20.1 ## 0.31.10 ### Patch Changes - c065064: fix: throw error for invalid response from o1 ## 0.31.9 ### Patch Changes - f2179f0: fix: upload test assets for error debugging ## 0.31.8 ### Patch Changes - 621c5f5: fix: spacing and format of browsing agent prompt ## 0.31.7 ### Patch Changes - 48dc5e3: fix: update prompt for update flow ## 0.31.6 ### Patch Changes - 52cd5ee: feat: add o1 support for browsing agent ## 0.31.5 ### Patch Changes - 7cafaf6: fix: add logline for test gen failure ## 0.31.4 ### Patch Changes - 0d8e839: fix: no response issues in o1 models ## 0.31.3 ### Patch Changes - cce6ced: feat: base changes for enabling o1 & using it for update using master agent scenario - Updated dependencies [cce6ced] - @empiricalrun/llm@0.9.3 ## 0.31.2 ### Patch Changes - Updated dependencies [267a65d] - @empiricalrun/r2-uploader@0.3.3 ## 0.31.1 ### Patch Changes - Updated dependencies [f317919] - @empiricalrun/r2-uploader@0.3.2 ## 0.31.0 ### Minor Changes - fd3af1f: feat: add support for test file marked serial ### Patch Changes - 6521cb0: fix: add support for method declaration ## 0.30.4 ### Patch Changes - Updated dependencies [684ddff] - Updated dependencies [684ddff] - @empiricalrun/reporter@0.20.0 ## 0.30.3 ### Patch Changes - 76d50f7: chore: add logs for file upload - 76d50f7: chore: add logs for file upload ## 0.30.2 ### Patch Changes - Updated dependencies [0541f38] - @empiricalrun/reporter@0.19.0 ## 0.30.1 ### Patch Changes - e07dd51: fix: remove redundant suites arg from stripAndPrependImports util - 69f3aa9: fix: add test for injectCodeSnippetBySuiteChain util ## 0.30.0 ### Minor Changes - c367dcb: feat: support nested new test-block creation via master agent ### Patch Changes - 56f3d2f: feat: support nested new test-block creation via coding agent ## 0.29.3 ### Patch Changes - c180953: fix: use test-case-id to get testCase detail - Updated dependencies [c180953] - @empiricalrun/reporter@0.18.5 ## 0.29.2 ### Patch Changes - Updated dependencies [fccb68a] - @empiricalrun/r2-uploader@0.3.1 ## 0.29.1 ### Patch Changes - Updated dependencies [62be3b2] - @empiricalrun/reporter@0.18.4 ## 0.29.0 ### Minor Changes - ad735ca: feat: send branch_name var as ci output ## 0.28.5 ### Patch Changes - Updated dependencies [da0362c] - @empiricalrun/reporter@0.18.3 ## 0.28.4 ### Patch Changes - 2d27074: fix: add test for nested test update ## 0.28.3 ### Patch Changes - 5ab2120: Revert "fix: incomplete llm response for ts error fix" ## 0.28.2 ### Patch Changes - 90dba16: feat: sanitize html attributes (e.g. style) to reduce input tokens - 178d285: fix: use suites information in coding agent for update flow ## 0.28.1 ### Patch Changes - a6fe36d: fix: incomplete llm response at the time of fixing ts errors ## 0.28.0 ### Minor Changes - a3761d0: feat: support test block extraction via suites param - 7ab94df: feat: add support for downloading latest build during test gen workflow ### Patch Changes - Updated dependencies [cc12707] - @empiricalrun/reporter@0.18.2 ## 0.27.10 ### Patch Changes - 66aa7b6: fix: master agent gives up on blank page and coding agent hallucination - Updated dependencies [aadad32] - Updated dependencies [bac164e] - @empiricalrun/reporter@0.18.1 - @empiricalrun/r2-uploader@0.3.0 ## 0.27.9 ### Patch Changes - 1017548: fix: revert single backslash handling ## 0.27.8 ### Patch Changes - 849ec6a: fix: trace and video not playing after generation ## 0.27.7 ### Patch Changes - ae6e28a: fix: handle different image input for different llms and handle escape characters in llm response - cf0a651: fix: master agent should generate code with correct page variable name - b1825cf: fix: handle typescript errors and add tests - Updated dependencies [ae6e28a] - @empiricalrun/llm@0.9.2 ## 0.27.6 ### Patch Changes - Updated dependencies [2f520d3] - @empiricalrun/r2-uploader@0.2.0 ## 0.27.5 ### Patch Changes - Updated dependencies [8bc7e9d] - @empiricalrun/reporter@0.18.0 ## 0.27.4 ### Patch Changes - Updated dependencies [2f0a135] - @empiricalrun/reporter@0.17.12 ## 0.27.3 ### Patch Changes - 2fc2f93: fix: handle consecutive action error handling - Updated dependencies [e7ae100] - @empiricalrun/reporter@0.17.11 ## 0.27.2 ### Patch Changes - 947dc1a: feat: support array types for testMatch and testIgnore in project detection ## 0.27.1 ### Patch Changes - ef6b74a: fix: stop button becomes start if page is reloaded ## 0.27.0 ### Minor Changes - 88067ed: fix: create test script remain in code after test gen error ## 0.26.0 ### Minor Changes - 5882a20: feat: add inline master agent support ### Patch Changes - 5882a20: fix: improve code agent accuracy with updated prompt ## 0.25.2 ### Patch Changes - Updated dependencies [ae142a5] - @empiricalrun/r2-uploader@0.1.3 ## 0.25.1 ### Patch Changes - Updated dependencies [7a51a93] - @empiricalrun/r2-uploader@0.1.2 ## 0.25.0 ### Minor Changes - a3fe47f: feat: deprecate browser agent and remove dependency of test object in create test ## 0.24.3 ### Patch Changes - Updated dependencies [be30850] - @empiricalrun/llm@0.9.1 ## 0.24.2 ### Patch Changes - ce74ef8: fix: local mock github flow ## 0.24.1 ### Patch Changes - b403cf9: fix: max token for llm generation ## 0.24.0 ### Minor Changes - 4b1a0cc: feat: support for code agent in update flow ## 0.23.14 ### Patch Changes - Updated dependencies [3027212] - @empiricalrun/llm@0.9.0 ## 0.23.13 ### Patch Changes - Updated dependencies [51b7e11] - @empiricalrun/r2-uploader@0.1.1 ## 0.23.12 ### Patch Changes - 6a1288a: feat: create r2 uploader package and support for retry - Updated dependencies [bff7e8b] - Updated dependencies [6a1288a] - @empiricalrun/reporter@0.17.10 - @empiricalrun/r2-uploader@0.1.0 ## 0.23.11 ### Patch Changes - 4c7b815: fix: playwright scripts missing post page navigation - Updated dependencies [a851891] - @empiricalrun/llm@0.8.3 ## 0.23.10 ### Patch Changes - 08bb613: chore: bump up the llm package version for testing - Updated dependencies [08bb613] - @empiricalrun/llm@0.8.2 - @empiricalrun/reporter@0.17.9 ## 0.23.9 ### Patch Changes - 6255f5f: chore: bump up the llm package version for testing - Updated dependencies [6255f5f] - @empiricalrun/llm@0.8.1 - @empiricalrun/reporter@0.17.8 ## 0.23.8 ### Patch Changes - Updated dependencies [1c4780e] - @empiricalrun/llm@0.8.0 ## 0.23.7 ### Patch Changes - 8946578: fix: incorrect langfuse trace being generated ## 0.23.6 ### Patch Changes - Updated dependencies [1bb7ce3] - @empiricalrun/llm@0.7.4 ## 0.23.5 ### Patch Changes - 6d954fd: feat: use custom playwright reporter - Updated dependencies [6d954fd] - @empiricalrun/reporter@0.17.7 ## 0.23.4 ### Patch Changes - d1068e6: fix: multiple session details call and missing generation id in requests - Updated dependencies [d1068e6] - @empiricalrun/reporter@0.17.6 ## 0.23.3 ### Patch Changes - Updated dependencies [f02f7a7] - @empiricalrun/llm@0.7.3 ## 0.23.2 ### Patch Changes - f6f488d: feat: add support for cancelling generation ## 0.23.1 ### Patch Changes - 108ee1b: fix: encode testGenToken before btoa ## 0.23.0 ### Minor Changes - 3b4b2f3: feat: update the dashboard ux with updated messages between dashboard and test-gen ## 0.22.12 ### Patch Changes - bfd6724: fix: avoid duplicate fixture imports ## 0.22.11 ### Patch Changes - 6771494: fix: handle dollar sign in generated code ## 0.22.10 ### Patch Changes - 5c43cd8: feat: support master agent to edit tests inside a describe block ## 0.22.9 ### Patch Changes - Updated dependencies [8a755ba] - @empiricalrun/reporter@0.17.5 ## 0.22.8 ### Patch Changes - Updated dependencies [9fa7d90] - @empiricalrun/reporter@0.17.4 ## 0.22.7 ### Patch Changes - Updated dependencies [5416dda] - @empiricalrun/reporter@0.17.3 ## 0.22.6 ### Patch Changes - Updated dependencies [b6a7231] - @empiricalrun/reporter@0.17.2 ## 0.22.5 ### Patch Changes - Updated dependencies [721c21c] - @empiricalrun/reporter@0.17.1 ## 0.22.4 ### Patch Changes - 2b9cf45: fix: text for agent being used - 1274236: fix: langfuse missing spans - Updated dependencies [1274236] - Updated dependencies [28d91a9] - @empiricalrun/llm@0.7.2 - @empiricalrun/reporter@0.17.0 ## 0.22.3 ### Patch Changes - Updated dependencies [c7c302f] - Updated dependencies [a84cb6d] - @empiricalrun/llm@0.7.1 ## 0.22.2 ### Patch Changes - d35239e: fix: update imports for test and expect to fixtures - Updated dependencies [2de783d] - @empiricalrun/llm@0.7.0 ## 0.22.1 ### Patch Changes - b96491f: fix: typescript errors for valid CommonJS module imports ## 0.22.0 ### Minor Changes - 4676ab7: feat: use gpt-4o-2024-08-06 as default ### Patch Changes - Updated dependencies [1770926] - Updated dependencies [4676ab7] - @empiricalrun/llm@0.6.0 ## 0.21.11 ### Patch Changes - e2b7a29: fix: codegen adding extra changes than the requested task - 29cad4a: feat: use vercel kv ## 0.21.10 ### Patch Changes - e077e2a: fix: verification agent should use model configured from test token ## 0.21.9 ### Patch Changes - Updated dependencies [2eb3598] - @empiricalrun/reporter@0.16.1 ## 0.21.8 ### Patch Changes - Updated dependencies [4642efd] - @empiricalrun/llm@0.5.0 ## 0.21.7 ### Patch Changes - Updated dependencies [7f2cc25] - @empiricalrun/reporter@0.16.0 ## 0.21.6 ### Patch Changes - Updated dependencies [ee7b03e] - @empiricalrun/reporter@0.15.0 ## 0.21.5 ### Patch Changes - 83d1382: fix: test code getting empty when using code gen - Updated dependencies [83d1382] - @empiricalrun/llm@0.4.5 ## 0.21.4 ### Patch Changes - Updated dependencies [09edb5f] - @empiricalrun/reporter@0.14.3 ## 0.21.3 ### Patch Changes - c92362f: fix: verification agent hallucination ## 0.21.2 ### Patch Changes - c3b31ae: fix: logic of extracting test block ## 0.21.1 ### Patch Changes - Updated dependencies [cc33982] - Updated dependencies [3649b78] - @empiricalrun/reporter@0.14.2 - @empiricalrun/llm@0.4.4 ## 0.21.0 ### Minor Changes - 3c714f7: feat: add master agent support ### Patch Changes - d7a24f0: fix: add lint rule for no floating promises - d1a1ff2: fix: restrict usage of llm basis token usage - Updated dependencies [3c714f7] - Updated dependencies [d1a1ff2] - @empiricalrun/llm@0.4.3 ## 0.20.5 ### Patch Changes - Updated dependencies [a9423b7] - @empiricalrun/llm@0.4.2 ## 0.20.4 ### Patch Changes - Updated dependencies [b2cace1] - @empiricalrun/reporter@0.14.1 ## 0.20.3 ### Patch Changes - Updated dependencies [701f1c9] - @empiricalrun/reporter@0.14.0 ## 0.20.2 ### Patch Changes - a1bdd4c: fix: missing messages in dashboard while creating test - Updated dependencies [a1bdd4c] - @empiricalrun/reporter@0.13.1 ## 0.20.1 ### Patch Changes - 6099fbe: fix: add a unit test for uploading test result and video - Updated dependencies [1647fbe] - @empiricalrun/reporter@0.13.0 ## 0.20.0 ### Minor Changes - 77a3e5a: feat: drop unused scenario and assert entities ### Patch Changes - Updated dependencies [bcf53ae] - @empiricalrun/llm@0.4.1 ## 0.19.1 ### Patch Changes - Updated dependencies [b7797f9] - @empiricalrun/reporter@0.12.3 ## 0.19.0 ### Minor Changes - 2087461: feat: remove LLMTracing class and use native methods ### Patch Changes - Updated dependencies [2087461] - @empiricalrun/llm@0.4.0 ## 0.18.0 ### Minor Changes - 4390c31: fix: filter video based on test name ## 0.17.6 ### Patch Changes - Updated dependencies [7b00d84] - @empiricalrun/reporter@0.12.2 ## 0.17.5 ### Patch Changes - Updated dependencies [5293f9e] - @empiricalrun/llm@0.3.0 ## 0.17.4 ### Patch Changes - Updated dependencies [61ee238] - @empiricalrun/reporter@0.12.1 ## 0.17.3 ### Patch Changes - Updated dependencies [8ef4bfa] - @empiricalrun/reporter@0.12.0 ## 0.17.2 ### Patch Changes - Updated dependencies [1c3fb96] - @empiricalrun/reporter@0.11.2 ## 0.17.1 ### Patch Changes - ac0256e: feat: remove support for update flag in test gen ## 0.17.0 ### Minor Changes - 914b207: fix: remove google sheet support in test gen ### Patch Changes - 74eefc4: fix: remove comments of browser agent ## 0.16.13 ### Patch Changes - fd6c8c5: fix: remove asserts from github description ## 0.16.12 ### Patch Changes - 9ad9089: fix: github output read file error ## 0.16.11 ### Patch Changes - 1daf179: fix: add outputs using core actions pkg ## 0.16.10 ### Patch Changes - ef600f4: chore: add github output logs ## 0.16.9 ### Patch Changes - 8b39547: chore: test sync api for writing to git output ## 0.16.8 ### Patch Changes - 1791ad0: chore: add logs for ci report ## 0.16.7 ### Patch Changes - be7898c: chore: add logs for ci report ## 0.16.6 ### Patch Changes - 52b7afc: feat: add tags to langfuse traces in code & browser agent ## 0.16.5 ### Patch Changes - Updated dependencies [0b41c5e] - @empiricalrun/reporter@0.11.1 ## 0.16.4 ### Patch Changes - 4c32cf9: feat: add support for uploading test video of browsing agent generated tests ## 0.16.3 ### Patch Changes - 8243b11: fix: add logs for errors while running playwright ## 0.16.2 ### Patch Changes - Updated dependencies [ffaa43d] - @empiricalrun/llm@0.2.0 ## 0.16.1 ### Patch Changes - 8b19502: fix: browsing agent message format ## 0.16.0 ### Minor Changes - af9724a: feat: add support for ai gateway ### Patch Changes - Updated dependencies [af9724a] - @empiricalrun/llm@0.1.0 ## 0.15.1 ### Patch Changes - 0727bad: fix: run browsing agent with correct project name matching the test case group ## 0.15.0 ### Minor Changes - 48720c0: feat: add support for reporter message to be available in dashboard ### Patch Changes - Updated dependencies [48720c0] - @empiricalrun/reporter@0.11.0 ## 0.14.2 ### Patch Changes - 92287cf: fix: replace has-text with contains pseudo for jquery ## 0.14.1 ### Patch Changes - ce5baaf: fix: filling action of browser agent ## 0.14.0 ### Minor Changes - f2eaa18: fix: support for css selectors like :contains ## 0.13.1 ### Patch Changes - 3ae0538: fix: langfuse flush events ## 0.13.0 ### Minor Changes - 35548e8: feat: add support for browsing agent for update scenario ## 0.12.0 ### Minor Changes - 67a4921: feat: drop support for yaml config support ## 0.11.1 ### Patch Changes - e8244a1: fix: specify project for browsing agent ## 0.11.0 ### Minor Changes - a1edabb: feat: support page reload as an action ### Patch Changes - 21dae78: fix: add sanitisation of browsing agent task ## 0.10.5 ### Patch Changes - 08ecca2: fix: test gen should quit after 3 consecutive errors and update dashboard sink message format - 08ecca2: fix: test gen should quit after 3 consecutive errors ## 0.10.4 ### Patch Changes - 5400dfa: fix: prepare files in subdirectory for browsing agent ## 0.10.3 ### Patch Changes - 08f12e9: fix: browsing agent output works with ci reporter ## 0.10.2 ### Patch Changes - 4c9f5a6: fix: inject pw locator script on page.goto - 5763c7e: fix: remove hasText filter from assert action ## 0.10.1 ### Patch Changes - dbacd41: chore: browsing agent should bubble up exit on error - 7b7d1fe: chore: remove redundant try-catch from browsing agent ## 0.10.0 ### Minor Changes - e29b407: feat: start with browsing agent ### Patch Changes - 057a0cb: feat: action for page navigation ## 0.9.0 ### Minor Changes - a6067de: feat: generate better playwright locators using playwright ## 0.8.2 ### Patch Changes - b8681d5: fix: update the action for locator ## 0.8.1 ### Patch Changes - a7e8c1c: fix: locator issues while generating tests ## 0.8.0 ### Minor Changes - 30f61bc: feat: add support for browsing agent ### Patch Changes - 71ffe00: feat: respect gitignore while building prompt context ## 0.7.3 ### Patch Changes - 9c66d6d: fix: empty steps in test case ## 0.7.2 ### Patch Changes - 9c3ee14: fix: import statements not getting detected correctly - 6ec6f56: fix: flag getting picked incorrectly ## 0.7.1 ### Patch Changes - 0aaf594: fix: make it easier to open trace ## 0.7.0 ### Minor Changes - f4a2519: feat: add support for json string support for adding / updating tests using test-gen ## 0.6.0 ### Minor Changes - 3e0e6d1: feat: add support for json string support for adding / updating tests using test-gen ## 0.5.1 ### Patch Changes - fb09f23: fix: build action in gh ## 0.5.0 ### Minor Changes - c09db7e: feat: add observability ### Patch Changes - 77dec1d: feat: add support for editing a scenario - 453c8f6: fix: ts feedbacks getting applied to other scenarios ## 0.4.2 ### Patch Changes - 2b3a59c: chore: more line breaks in ci report ## 0.4.1 ### Patch Changes - 51294b4: fix: indentation of the ci summary ## 0.4.0 ### Minor Changes - f8b58d4: feat: add support for github ci reporter ## 0.3.0 ### Minor Changes - f993c4c: feat: add support for creating test case using google sheet ### Patch Changes - f993c4c: fix: decoding error of new line in evnironment variable ## 0.2.0 ### Minor Changes - 5333d3b: feat: add support for creating test case using google sheet ## 0.1.3 ### Patch Changes - b32c561: fix: increase readability of logs ## 0.1.2 ### Patch Changes - bf0228b: fix: running test-gen throwing package / module not found error ## 0.1.1 ### Patch Changes - c75a102: chore: add publish package support ## 0.1.0 ### Minor Changes - 20c6004: feat: add support for package publish